Our Recent Salon
We had a Salon a couple of weeks ago that was pretty exciting for us as a new group. Four of our teammates were able to attend in person, while our two legal eagles helped us ready for the event. More than 40 people showed up to participate, and the conversations and interest were significant.
I was able to leverage Athropic’s brand new Claude 2 large language model to summarize the recording of the event (thanks Dru!) and pull together highlights. Please note, I’ve edited the results slightly, but Claude 2 took a PDF and summarized it in seconds, allowing me to blog from a draft, vs. a blank page (the biggest drawback of blogging regularly!).
Highlights:
AI Governance Group (AIGG) formed to help businesses adopt AI tools responsibly
Social media and internet created an environment of mistrust and misinformation - truth has been turned upside down in the 20 years since the conversational web took hold.
We’re glad we’re exploring and talking about the implications - especially ethically - of this tech transformation
Need policies in place for employee handbooks to protect companies, employees, their IP and brands using AI
Bias is a major concern with AI systems reflecting existing biases in data - which are continually being ‘discovered’
Ownership and IP with generative AI like chatGPT is an open question - we had lots of discussion around this
Brady Benware, GPT Now, provided real-world technology examples of where clients are exploring the use of AI
Lots of conversations around hiring and firing using AI assistant tools - and the need (overall) for human interventions
The audience questions touched on both the positive potential of AI, like for climate change, as well as risks around bias, copyright, and unintended consequences. There was interest in both practical applications of AI as well as broader societal impacts and how it should be regulated. Our panelists provided perspectives from technical, legal, creative and business standpoints.
Major Questions:
What policies are needed for using AI?
A: Employee handbook policies (including moonlighting), privacy policies, terms of use
How can AI be used for good?
A: Climate change data analysis, medical research, freedom to be more strategic and creative
Who owns the IP with generative AI?
A: Unclear currently, the copyright office is in an “it depends” mode right now, but it’s clear that human creativity is essential to obtaining copyrights
How can AI help address climate change?
A: Analyzing large datasets, influencing public behavior
How can AI be used without bias?
A: Careful data selection, human oversight of results. The data is key - no doubt. We’re in early days, and still discovering (especially) how many of these tools are trained.
In summary, the panel discussed both the risks and opportunities of AI and emphasized the need for thoughtful governance and oversight to harness its potential for social good. Thoughtful human oversight is critical, along with an attention to detailed data source understanding…
Overall it was a thought-provoking discussion about both the promise and pitfalls of AI, and we definitely plan to do it again!
Thank you to Susan Towers of Sandymount Search for being such a great moderator and hostess. And to the folks at GRAYBOX for allowing us to use the space for the gathering.
